* [PentAGI v2.1.0](https://github.com/vxcontrol/pentagi) – Autonomous AI-driven penetration testing platform that runs sandboxed attacks, orchestrates specialized agents, and produces detailed vulnerability reports.

* [Codex Autoresearch v2.0.0](https://github.com/TheGreenCedar/codex-autoresearch) – Plugin enabling measured experiment loops to iteratively optimize repo-local benchmarks and produce reviewable changes.

* [Cordum v1.0.0](https://github.com/cordum-io/cordum) – Deterministic control plane for autonomous AI agents and external workers, using NATS, Redis, and policy-first workflow orchestration.

* [Strands Agents - TypeScript SDK v1.3.0](https://github.com/strands-agents/sdk-typescript) – Model-driven TypeScript/JavaScript SDK for building type-safe, extensible AI agents with streaming and multiple model provider support.

* [AI Maestro v0.35.0](https://github.com/23blocks-OS/ai-maestro) – Agent orchestrator providing persistent memory, agent-to-agent messaging, and a dashboard to manage agents across multiple machines.

* [WUPHF v0.186.0](https://github.com/nex-crm/wuphf) – Slack-like collaborative office where autonomous AI agents share memory and coordinate to run tasks continuously.

* [Project Golem v9.6.14](https://github.com/Arvincreator/project-golem) – Browser-in-the-loop multiagent AI using Web Gemini and Node.js for web automation and social platform integration.

* [OpenSwarm v0.4.4](https://github.com/unohee/OpenSwarm) – Autonomous orchestrator that runs Claude Code agents to process Linear issues, produce code changes, report to Discord, and store memory in LanceDB embeddings.

* [Milady v2.0.0-alpha.140](https://github.com/milady-ai/milady) – Local privacy-focused AI assistant that runs on the user's computer and manages agents, sessions, and messaging integrations.